Little information is available on staged Gamma Knife surgery (GKS) with an interval of 3 years or more when used to treat arteriovenous malformations (AVMs) with volumes larger than 10 cm(3). The goal of this study was to increase knowledge in this area by reporting the authors' experience. The authors describe an institutional review board-approved retrospective study in which they examined databases including information on 250 patients who consecutively underwent GKS for cerebral AVMs during a 16-year period (1988-2004). Among the 250 patients the authors identified 31 patients (12.4%, 15 female and 16 male patients with a mean age of 29 years [range 10-63 years]) in whom 2-stage GKS was intentionally planned at the time of initial treatment because the volume of the AVM nidus was larger than 10 cm(3). The most common presentation was bleeding (14 patients), followed by seizures (9 patients), incidental findings (7 patients), and headache with scintillation (1 patient). One patient underwent GKS for the treatment of 2 AVMs simultaneously, and thus 32 AVMs are included in this study. The mean nidus volume was 16.2 cm(3) (maximum 55.8 cm(3)). In all 31 patients, relatively low radiation doses (12-16 Gy directed at the periphery of the lesion) were intentionally used for the first GKS. The second GKS was scheduled for at least 36 months after the first. Complete nidus obliteration was obtained after the first GKS in 1 patient. To date, 26 patients have undergone a second procedure with a post-GKS mean interval of 41 months (range 24-83 months); 2 other patients refused to undergo the second GKS, and no further treatment was given because of severe morbidity in 1 case and death due to bleeding in the other case. Among the 26 patients who did undergo a second procedure, 3 patients refused follow-up digital subtraction (DS) angiography, another is scheduled for follow-up DS angiography, and 2 patients died, one of bleeding and the other of an unknown cause. The remaining 20 patients underwent follow-up DS angiography. Complete nidus obliteration was confirmed in 13 patients (65.0%) and remarkable nidus shrinkage in the other 7 patients (35.0%). In 2 of these 7 patients, a third GKS achieved complete nidus obliteration. Therefore, the cumulative complete obliteration rate in this series was 76.2% (16 of 21 eligible patients). Seven patients (22.6%) experienced bleeding. The bleeding rates were 9.7%, 16.1%, 16.1%, and 26.1%, respectively, at 1, 2, 5, and 10 years post-GKS. There were 2 deaths and 3 cases of morbidity (persistent coma, mild hemimotor weakness, and hemianopsia in 1 patient each). Hemorrhage did not produce neurological deficits in the other 2 patients. During the mean post-GKS follow-up period of 105 months (range 42-229 months) to date, mild symptomatic GKS-related complications occurred in 2 patients (6.5%); these were classified as Radiation Oncology Group Neurotoxicity Grade 1 in 1 patient and Grade 2 in the other. Among various pre-GKS clinical factors, univariate analysis showed only patient age to impact complications (hazard ratio 0.675, 95% CI 0.306-0.942, p = 0.0085). The rate of complications in the pediatric cases was 33.3%, whereas that in the adolescent and adult cases was 0% (p = 0.0323). Although a final conclusion awaits further studies and patient follow-up, these results suggest 2-stage GKS to be beneficial even for relatively large AVMs.